Origin of solar systems 30 June - 2 July 2009 by Klaus Jockers Max-Planck-Institut of Solar System Science Katlenburg-Lindau Part 5 Condensation and growth of solid bodies in protoplanetary disks

Theoretical considerations concerning protoplanetary disks: Why is the disk expected to be tapered (scale height versus distance from host star) Angular momentum in the disk Condensation and growth of solid bodies Time scales of planetesimal formation Infall of grains onto the disk Growth of sub-meter particles by coagulation The drift problem in a disk partially supported by pressure Growth of planetesimals > 1 km (gravitational regime) Outline

The extent of a Keplerian disk perpendicular to the disk plane: Pringle, J.E., accretion disks in astrophysics, Ann. Rev. Astron. Astrophys. 1981, 19, ) Keplerian disk: Mass of disk negligible as compared to mass of central star. Assumption: no forces except gravitation. Hydrostatic equilibrium perpendicular to the disk plane: Replace the density ρ by the pressure p using the ideal gas law: Integrate, separating the variables: Introduce Gaussian scale height H: Note z 2, not z! By comparison: For physically reasonable temperature distributions, like T~r -1/2, H rises with distance from the central star, i.e. the disk is tapered.

Theoretical considerations concerning protoplanetary disks: Why is the disk expected to be tapered (scale height versus distance from host star) Angular momentum in the disk Condensation and growth of solid bodies Time scales of planetesimal formation Infall of grains onto the disk Growth of sub-meter particles by coagulation The drift problem in a disk partially supported by pressure Growth of planetesimals > 1 km (gravitational regime) Outline

Internal dynamical evolution of the disk: Redistribution of angular momentum can provide additional mass to the central star. Magnetic torque can reduce rotation of star if ionization is high (frozen-in magn. field). Process needs ionized gas that may not be available at large distances from the proto- Sun. Protoplanetary disks apparently do not extend all the way down to the surface of the star. Magnetic interactions at the corotation point funnel some of the disk’s gas onto the star and expel other gas in rapid centrifugally driven bipolar outflow which carries with it a substantial amount of angular momentum (?). Gravitational torques: Nonaxisymmetric local instabilities can create spiral density waves like in galaxies or in the Saturnian ring system which limit the allowed mass of the disk. Large protoplanets may clear annual gaps surrounding their orbits, excite density waves transporting angular momentum outwards. Similar effects can arise if protostar rotates sufficiently rapidly to become triaxial. But as observations indicate, such high rotation rates seem to be unlikely.

Internal dynamical evolution of the disk (continued): Viscous torques: Molecules move on Keplerian orbits, i. e. its transverse speed decreases outwards. Collisions (or turbulence) transfer mass inward and angular momentum outward (Lynden-Bell and Pringle MNRAS 168, , 1974). Evolution on diffusion timescale: l = radius of the disk. Viscosity largely unknown. Molecular viscosity: l fp mean free path of molecules and c s sound speed. Turbulent viscosity: < α ν < H z depends on temperature and the disk temperature on opacity perp. to midplane. Well inside the orbit of Mercury the interstellar dust grains are all evaporated and opacity is caused by H 2 O and CO molecules and H ionization. At larger distances from the star the temperature is below 2000K and micrometer-sized dust is the dominant source of opacity.

Theoretical considerations concerning protoplanetary disks: Why is the disk expected to be tapered (scale height versus distance from host star) Angular momentum in the disk Condensation and growth of solid bodies Time scales of planetesimal formation Infall of grains onto the disk Growth of sub-meter particles by coagulation The drift problem in a disk partially supported by pressure Growth of planetesimals > 1 km (gravitational regime) Outline

Condensation and growth of solid bodies (Imke de Pater, Jack J. Lissauer: Planetary Sciences, Cambridge U. Press, Jack J. Lissauer, Planet Formation, Ann. Rev. Astron. Astrophys. 1993, 31: ) Timescales for planetesimal formation: The age of most chondrites (primitive meteorites) is 4.56 Gyr and they formed within a period ≤ 20 Myr Evidence from extinct 26 Al (t 1/2 = 0.72 Myr) in carbonaceous chondrites suggests that first solid material formed only a few million years after the last injection of freshly nucleosynthesized matter (but other explanations exist). This timescale is the timescale of collapse discussed earlier. Evidence is based on observation of the daughter product 26 Mg close to 27 Al. The freshly nucleosynthesized matter could come from stellar winds produced by a nearby asymptotic giant branch (AGB) star. Isotope analysis of primitive meteorites indicates that they still contain interstellar grains.

Theoretical considerations concerning protoplanetary disks: Why is the disk expected to be tapered (scale height versus distance from host star) Angular momentum in the disk Condensation and growth of solid bodies Time scales of planetesimal formation Infall of grains onto the disk Growth of sub-meter particles by coagulation The drift problem in a disk partially supported by pressure Growth of planetesimals > 1 km (gravitational regime) Outline

Protoplanetary disk: Infall stage Gas and Dust Duration of infall stage comparable to free-fall collapse time of the core ~ yr. Matter with low specific angular momentum falls into the central star. Matter with high specific angular momentum cannot reach the central star and forms the disk. In the following it is shown that for such matter exactly half of the gravitational energy gained during infall goes into the kinetic energy of the orbiting body and the other half is converted to heat. Kepler velocity: Body (mass m) on circular orbit around central star with mass M: r disk plane z

Parcels of gas fall from both sides of the disk from infinity to a circular orbit at heliocentric distance r SUN and meet there. Gravitational energy gained: Kinetic energy: → Half of gravitational energy is converted to orbital kinetic energy, the other half per unit mass, is available for heat. Equate half of gravitational energy with thermal energy per particle and find: At 1 AU and 1 solar mass v c = 30 km s-1 and the temperature in a hydrogen gas ~7 x 10 4 K. But temperature falls quickly because of radiative cooling. When two clouds meet from both sides of the forming disk, shock fronts form with temperatures ~1500 K at 1 AU and ~100 K at 10 AU. z r disk plane z Accretion

Infall of grains onto the disk: Acceleration of grains toward disk: ρ g density of gas, ρ grain density, c s local sound speed, R grain radius. At 1 AU T = K, ρ g = g cm -3, c s = 2.5x10 5 cm s -1, v z = 0.03 (z/H z ) cm s -1. (H 2 nebula). For 1 μm grain with ρ=1 gcm -3 it takes 10 6 years to fall halfway toward midplane, 10 7 years for 99.9% of the distance. Coagulation is needed to form the disk in the available time. n is Keplerian orbital angular velocity. (mean motion) Equilibrium settling speed: n = v Kep /r Epstein drag

Theoretical considerations concerning protoplanetary disks: Why is the disk expected to be tapered (scale height versus distance from host star) Angular momentum in the disk Condensation and growth of solid bodies Time scales of planetesimal formation Infall of grains onto the disk Growth of sub-meter particles by coagulation The drift problem in a disk partially supported by pressure Growth of planetesimals > 1 km (gravitational regime) Outline

Coagulation is different for fluffy and smooth particles. It depends on electric charge, electric conductivity in the grain and molecular forces. Presently experiments are performed in space under microgravity and in the laboratory. Planetesimal formation starts with the growth of fractal dust aggregates, followed by compaction processes. As the dust-aggregate sizes increase, the mean collision velocity also increases, leading to the stalling of the growth and possibly to fragmentation, once the dust aggregates have reached decimeter sizes. Current models indicate a settling time into mm-sized bodies in 10 4 years. For more details see: Jürgen Blum, Dust agglomeration, Advances in Physics 2006, Vol. 55, 881–947. Jürgen Blum, Gerhard Wurm, The growth mechanisms of macroscopic bodies in protoplanetary disks, 2008, Ann. Rev. Astron. Astrophys. 46: Coagulation

Theoretical considerations concerning protoplanetary disks: Why is the disk expected to be tapered (scale height versus distance from host star) Angular momentum in the disk Condensation and growth of solid bodies Time scales of planetesimal formation Infall of grains onto the disk Growth of sub-meter particles by coagulation The drift problem in a disk partially supported by pressure Growth of planetesimals > 1 km (gravitational regime) Outline

The problem of inward drift in a partially pressure supported disk: Gas circles star slightly less rapidly than Keplerian rate. Effective gravity felt by gas: In circular orbits, the effective gravity is balanced by centrifugal forces r Sun n 2. Since the pressure is much smaller than gravity we can approximate the angular velocity n gas as For estimated protoplanetray disk parameters the gas rotates 0.5% slower than the Keplerian speed. But large particles must move with Keplerian speed, otherwise they will fall into the protostar!

Radial inward drift of planetesimals: Particles moving at (nearly) the Keplerian speed encounter a headwind which removes part of their orbital angular momentum and causes them to spiral inward towards the star. Small particles are strongly coupled to the gas and therefore drift very slowly. Kilometer-sized bodies also drift inwards very slowly, because their surface to mass ratio is small. Peak inward drift rates occur for particles that collide with roughly their own mass within one orbital period. At 1 AU a meter-sized body drifts inwards at the fastest rate ~10 6 km yr -1. Because of the difference in (both radial and azimuthal) velocities, small (subcentimeter) grains can be swept on by larger grains which in turn move toward proto-Sun.

Particles must grow through this size range quickly, otherwise they will be lost. ρ dust = 0.5 g cm g cm g cm -3 No solution of this problem exists at present.

Theoretical considerations concerning protoplanetary disks: Why is the disk expected to be tapered (scale height versus distance from host star) Angular momentum in the disk Condensation and growth of solid bodies Time scales of planetesimal formation Infall of grains onto the disk Growth of sub-meter particles by coagulation The drift problem in a disk partially supported by pressure Growth of planetesimals > 1 km (gravitational regime) Outline

Growth from planetesimals to planetary embryos, gravitational regime: For bodies > 1 km major forces are gravitational interaction and physical collisions and gas drag. Collision between planetesimals: v relative speed at large distances v e escape speed Impact velocity v i ≥ v e v i ≥ 6 m s -1 for rocky 10 km body. Some of the kinetic energy of the colliding particle must be dissipated → Rebound velocity = v i ε with ε ≤ 1. If v i ε ≤ v e particle accretes sooner or later. The relative speed between planetesimals is critical: Only if v « v e probability for capture of the particle high. If relative speed is too high, small grains will not accrete on large grains → instead sandblasting of growing planetesimals.

At a larger scale, growth from cm-sized to kilometer sized planetesimals depends primarily on the relative motions between the various bodies. Model calculations Note: Small relative speeds for small grains. Reduced relative speeds for large masses of similar size. Plateau for relative speeds of small and large bodies.

Growing planets in the gravitational regime: Without proof: Gravitational enhancement factor Mass accretion, ρ s volume density of planetesimal swarm: R radius of planetary embryo. If we express v e by the radius R of the planetesimal, we get: and i.e. for v ≈ v e the mass grows ~ R 2 and for v « v e the mass grows ~ R 4 (runaway growth). Safronov, V.S.: Evolution of the proto-planetary cloud and formation of the Earth and planets, Moskva, Nauka, 1969.

Growing planets in the gravitational regime (continued): If one body is larger than all the others, it will not stir up the mean relative velocity v as much as if all bodies have similar size. This will allow continued fast growth until three-body interactions become important.

Transfer to surface densities and calculation of planetesimal growth in radius: Gaussian scale height (calculated earlier) n is mean motion. Lissauer: If the proto-Sun’s gravity is dominant force in vertical direction and if the relative velocity between planetesimals is isotropic, then The two scale heights are the same except of. Transfer to surface mass density of planetesimals σ ρ (g cm -2 ): Growth of radius: ρ p density of planetary embryo

Growth time of planets: For Earth F g = 7, σ ρ =10 g cm -2, n = 2x10 -7 s -1, ρ p = 4.5 g cm -3, growth time 2x10 7 yr, or better 10 8 yr, if depletion of planetesimals in later stage of accretion is considered. Problems with outer planets. For Jupiter σ ρ = 3 g cm -2, heavy element mass Earth masses, growth time > 10 8 yr. Surface density of solar nebula drops ~ r -3/2, growth time of Neptune is many times the solar system age.

Hill Sphere

End of growth of planetary embryos Area within reach of the growing embryo is ~4 times its Hill sphere. Hill sphere: sphere of gravitational influence (limited by Lagrange points, previous view graph). Radius R H of Hill sphere: Mass of planetary embryo which has accreted all mass within a ring of width 2Δrסּ: If Δrסּ = 4 R H we obtain maximum mass M i (in g) to be accreted by a planetary embryo orbiting a star of 1 Mסּ: For Earth M i = g. 1 Earth mass = g.

Making planetary embryos close to the Earth, numerical calculations: see Eiichiro Kokubo, Planetary accretion: From Planetesimals to Protoplanets, Rev. Mod. Astronomy 14, , 2001.

Final stages of planetesimal accumulation The self-limitating nature of runaway growth implies that massive protoplanetary embryos form at regular intervals in semimajor axis. Their random velocities are no longer strongly damped by energy equipartition with the smaller planetesimals. Therefore the embryos will pump up each other’s velocities. The orbital excentricities will increase and the orbits possibly intersect. Subsequent orbital evolution is governed by close gravitational encounters and violent, highly inelastic collisions. (The Earth’s moon is believed to be generated during such a collision). The orbital evolution in the inner solar system can be studied with programs calculation the motion of N bodies in the gravitational field of the Sun. During this process of mutual violent collisions the chemical composition of the resulting planets is averaged over some range of heliocentric distance.

Summary, part 1 In a Keplerian disk the Gaussian pressure scale height = For reasonable temperature profiles like T~r -1/2, H rises with distance from the central star Because of the need to conserve angular momentum, gas and dust do not fall directly on the protostar, but fall parallel to the momentum vector into the disk. The turbulent disk transports mass inward and momentum outward and in this way allows accretion onto the protostar. For the solar system the time of planetesimal formation is known fairly accurately from the evidence of extinct 26 Al (half life 0.72 Myr) in primitive meteorites. Micron-sized particles can coagulate and grow to millimeter size in ~10 4 years. Growth beyond meter size is hindered as particles of this size experience strong gas drag and may therefore be swept into the host star in ~100 years.

Summary, part 2 Km-sized planetesimals have sufficient gravity to grow by gravitationally attracting other bodies. For v ≈ v e (escape velocity) dM/dt ~ R 2, and for v « v e dM/dt ~ R 4 (runaway growth). The Earth can be made in – 10 8 years, but, in accordance with Chapter 4 of this lecture, the gaseous planets cannot be formed by accretion of solid planetesimals alone. Once massive terrestrial protoplanets have formed, they will pump up each other’s velocities. Violent collisions may occur until the planets have at last reached orbits that are stable for billions of years.